Bringing NFL Data To Life

“The NFL’s inaugural Big Data Bowl is here. This sports analytics contest from NFL Football Operations is looking for talented members of the analytics community — from college students to professionals — to contribute to the NFL’s continuing evolution in the use of advanced analytics.”

https://operations.nfl.com/the-game/big-data-bowl/

When I saw this website I was in a dream. Unfortunately, I was about a month late in finding this article. I did manage to get signed up and get access to their GitHub repo with a ton of data. It was Christmas all over again. Anyone that knows me understands this is where I get my “Coach Fred” title. I’ve worked in different facets of technology for twenty plus years. I’ve been coaching youth football for about ten years. Both have brought much joy to my life. I read Moneyball four years ago and started applying the same concepts to youth football. Clearly, this has definitely given me an advantage over a lot of teams. I can’t tell you how much of an advantage computer vision has given me over other teams. I coach at a park that doesn’t get the highly talented athletes, but the kids tend to overachieve.

Football Is A Game Of Inches

Not only is it a game of inches. It’s a game of data, points, slopes, angles, positions, x/y axis, and all kinds of great math. What happens when you combine video and data with python, R, tensorflow, opencv, kinovea, mac, linux, Digital Ocean, Microsoft, and AWS. You can create a powerful coaching system. A system that as long as the players can follow the rules, they will win the game. I know no one likes to hear about the Patriots, but it is a system.

How To Build Your First Advanced Coaching System

First, let’s download RStudio

I’ve downloaded it on Windows, Linux, and Mac. By far, Linux did require some extra libraries to be downloaded. Once you have the screen above running make sure you at least running version 3.5.2, otherwise you will have problems with libraries. You can copy and paste this right in the left pane in RStudio:

install.packages(‘gganimate’)

install.packages(‘tidyverse’)

install.packages(‘cowplot’)

install.packages(‘gifski’)

install.packages(‘png’)

Once these are installed you can add this first part of the code.

The following code will read in the data the results from this play:

If you get errors, it’s probably due to a library not get installed correctly. You’ll have to go back and check through the errors. Now here comes the really awesome part, animating the data.

And here is the outcome after entering the above code:

More To Come

Using the data is going to give the teams who use it correctly and edge in the game. If coaches learn to use the data, more games will be won through analytics. Now, the players still have to do their part, but when you have a cheat sheet (I should be careful how I say “cheat”) there’s no limit to your team's success.